Viruses that specifically replicate and kill tumor cells are highly attractive and novel cancer therapies. After an initial infection, these viruses replicate and spread within the tumor, killing cancer cells while leaving normal cells unharmed. We have developed several viruses that are in clinical trials for patients with incurable tumors. These viruses exhibit different cell killing properties and replication kinetics. Vesicular stomatitis virus (VSV) kills cells rapidly by lysis and spreads from cell to cell by diffusion. In contrast, viruses derived from the Edmonston vaccine strain of measles virus (MV) spread by formation of syncytia. The outcome of tumor therapy with viruses is highly complex since it depends on the dynamic interactions among various populations including tumor cells, the virus, and various cellular components of the immune response and the local architecture of the tumor that may hinder or enable spread of the infection. Understanding the outcomes of such therapy requires a rigorous analysis of the dynamics of therapy in vivo. This necessitates (i) methodologies that enable serial and non-invasive quantification of the tumor cell population and virus load and (ii) mathematical and computational tools to understand these dynamics. We have developed replication competent VSV as well as MV derivatives that express the thyroidal sodium iodide symporter (NIS) in cells. Infected tumor cells express NIS and concentrate radioactive isotopes that can be imaged with single photon emission computerized tomography enabling us to serially and non- invasively monitor the biodistribution and propagation of these virus infected cells in tumors. Simultaneously, we have developed mathematical and computational models that capture the essential components of the dynamic interactions between viruses and their target tumors. In this grant application, we plan to: (i) establish the relationship between NIS mediated isotope uptake and intratumoral virus replication in order to be able to determine the absolute population of virus infected cells in the tumor at any time after the start of therapy. This analysis will look at various tumors to accommodate the expected inter tumor variability in NIS expression, isotope concentration as well as amplification and spread of the virus. We will also determine the impact of the initial distribution of virus within the tumor on its spread and the level of tumor control. (ii) We will determine how the architecture of tumors varies across a number of cancer models and attempt to understand how this architecture interacts with the oncolytic virus and influences the therapeutic results. (iii) We will continue to develop more appropriate mathematical and computational models that not only capture the in vivo dynamics but also are able to make meaningful predictions that will be tested in relevant animal models and allow us to optimize therapy. Our in vivo and in silico studies will be statistically correlated using novel approaches to determine the 'simplest'in silico model that reliably captures the dynamics of therapy and makes meaningful predictions that can be experimentally tested. We hypothesize that understanding the complex dynamics of tumor virotherapy will eventually help us with faster and safer translation of these novel therapies into the clinic and be associated with considerable savings by guiding the design of the relevant animal experiments and clinical trial design.